3D Reconstruction of Skin Surface from Image Sequence

This paper proposes a new method for reconstruction a shape of skin surface replica from shading image sequence taken with di erent light source directions. Since the shading images include shadows caused by surface height uctuation, and specular and inter re ections, the conventional photometric stereo method is not suitable for reconstructing its surface accurately. In the proposed method, we choose intensity data which do not include specular and inter re ections and self-shadows so that we can calculate accurate normal vector from the selected intensity data using SVD (Singular Value Decomposition) method. The experimental results from real images demonstrate that proposed method is e ective for shape reconstruction from shading images which include specular and inter re ections and self-shadows.